About

Bárbara Barros Carlos

Hello, I am Bárbara! I work at the intersection of numerical optimization, control, and estimation for robotics — mostly Nonlinear Model Predictive Control (NMPC) and Moving Horizon Estimation (MHE), with a focus on getting them running in real time on embedded hardware. I've worked on embedded numerical optimization for quadrotors at the Systems Control and Optimization (syscop) Laboratory in Freiburg, Germany, and on embedded NMPC and optimization-based shared control at the Robotics and InteractionS (RIS) team at LAAS-CNRS in Toulouse, France.

I like bridging theory and practice — taking advanced control and estimation algorithms out of the paper and into safety-critical, dynamic environments.

Experience

• Parrot — State Estimation & Sensor Fusion Engineer
  Paris, France · Jul 2026 – Present

  • Improving GNSS-denied navigation robustness for state estimation on autonomous drones
  • Enhancing sensor fusion algorithms combining IMU, GPS, and visual odometry for improved accuracy

• Stanley Robotics — Trajectory Planning, Control & State Estimation Engineer
  Paris, France · May 2022 – Jun 2026

  • Led development of real-time algorithms for the motion generation of Stan

• ITIC — Researcher
  Fortaleza, Brazil · Mar 2015 – Jun 2017

  • Quadrotor hardware technician
  • Developed an autopilot using BeagleBone Black
  • Technical project writer

• LIT/IFCE — Undergraduate Research Assistant
  Fortaleza, Brazil · Aug 2014 – Aug 2017

  • Developed a quadrotor prototype

• ITIC — Education Robotics Teacher
  Fortaleza, Brazil · Dec 2013 – Dec 2014

  • PIC 18FXX5X-based embedded systems development
  • Taught logic, programming, electricity, and basic electronics

Education

• Sapienza Università di Roma — Postdoctoral Research Fellow
  Rome, Italy · May 2021 – May 2022
  Line of research:

  • NMPC for unmanned aerial vehicles (UAVs)

• Sapienza Università di Roma — PhD in Automatica, Bioingegneria e Ricerca Operativa
  Rome, Italy · Nov 2017 – Jul 2021
  Thesis:

  • Real-Time Nonlinear Model Predictive Control for Motion Generation in Robotics Systems

• Laboratoire d'analyse et d'architecture des systèmes (LAAS - CNRS) — Visiting PhD Student
  Toulouse, France · Oct 2019 – Mar 2020
  Lines of research:

  • Embedded NMPC applied to quadrotors
  • Human-quadrotor mixed-initiative control via real-time NMPC
  • NMPC for quadrotor-slung payload system for the task of transportation on suspension

• IMTEK, University of Freiburg — Visiting PhD Student
  Freiburg, Germany · May 2019 – Oct 2019
  Lines of research:

  • Least conservative constraint formulation for real-time motion generation of the Crazyflie nanoquadrotor
  • Sensitivity-based real-time NMPC applied to quadrotors
  • NMPC for periodic motion generation and orbital stabilization of a double-inverted pendulum
  • Gain expertise in embedded numerical optimization methods

• Instituto Federal do Ceará — BSc in Mechatronics Engineering
  Fortaleza, Brazil · Aug 2013 – Jun 2017
  Thesis:

  • Modeling, Control and Simulation of a Quadrotor for Attitude Stabilization

Publications

• Conference Papers
  Real-Time NMPC for an Automated Valet Parking with Loaded-Based Safety Constraints and a Path-Parametric Model
  B. Barros Carlos, M. Williams, and B. Pelourdeau — International Conference on Intelligent Robots and Systems (IROS), 2023

  An Efficient Real-Time NMPC for Quadrotor Position Control under Communication Time-Delay
  B. Barros Carlos, T. Sartor, A. Zanelli, G. Frison, W. Burgard, M. Diehl, and G. Oriolo — International Conference on Control, Automation, Robotics and Vision (ICARCV), 2020

• Journal Articles
  Towards Safe Human-Quadrotor Interaction: Mixed-Initiative Control via Real-Time NMPC
  B. Barros Carlos, A. Franchi, and G. Oriolo — IEEE Robotics and Automation Letters (RA-L), 2020

  Least Conservative Linearized Constraint Formulation for Real-Time Motion Generation
  B. Barros Carlos, T. Sartor, A. Zanelli, M. Diehl, and G. Oriolo — IFAC-PaperOnLine, 2020

  Enforcing Constraints over Learned Policies via Nonlinear MPC: Application to the Pendubot
  G. Turrisi, B. Barros Carlos, M. Cefalo, V. Modugno, L. Lanari, and G. Oriolo — IFAC-PaperOnLine, 2020

Mentoring

• Stanley Robotics — Internship & MSc Thesis Supervisor
  Paris, France · 2025
  

  • NMPC framework for safe and robust navigation of Stan in narrow lanes under unmeasured disturbances

• Sapienza Università di Roma — MSc Thesis Supervisor
  Rome, Italy · 2021
  

  • Distributed NMPC for tracking and formation control of multi-UAV systems in forest-like environments

• Sapienza Università di Roma — Project Supervisor
  Rome, Italy · 2018
  

  • Flying inverted pendulum, supervised for a 1st-year student of Scuola Superiore di Studi Avanzati Sapienza (SSAS)
  • Nonlinear observability and estimation via the dynamic bearing observability matrix for multi-agent systems, supervised for the Control of Autonomous Multi-Agent Systems course

Community

• Efficient Real-Time NMPC for the Crazyflie 2.1 nanoquadrotor
  Open-source ROS stack · 2019
  
  • Modular, real-time NMPC implementation tailored to the Crazyflie 2.1, with time-delay compensation and actuator bound enforcement
  • Built on acados, leveraging its real-time iteration SQP scheme with Gauss-Newton Hessian approximation
  • Quadratic subproblems solved via HPIPM (interior-point method) on top of BLASFEO, with partial condensing applied to further reduce solution times

Skills